adr

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill makes runtime HTTP calls to the Confluence REST API (e.g. https://${cfg.site}/wiki/rest/api/... and https://${cfg.site}/wiki/api/v2/...) to fetch page bodies that are then injected into generated prompts/sub-agent tasks (e.g., the review task), so external content directly controls the agent's prompts and is required for the skill to function.